Coronavirus Disease 2019 and Prevalence of Chronic Liver Disease

A Meta-analysis

Alessandro Mantovani; Giorgia Beatrice; Andrea Dalbeni

Disclosures

Liver International. 2020;40(6):1316-1320. 

At present, there is scarce information regarding the global prevalence of chronic liver disease in individuals with coronavirus disease 2019 (COVID-19) disease, which is becoming a global pandemic. The aim of this study was to assess the overall prevalence of chronic liver disease among patients with COVID-19 disease by meta-analysing data in observational studies and to investigate the relationship between liver damage and COVID-19 disease. We included 11 observational studies for a total of 2034 adult individuals (median age 49 years [IQR 45–54], 57.2% men). The overall prevalence of chronic liver disease at baseline was 3% (95% CI 2%-4%; I 2 = 29.1%). Individuals with severe COVID-19 disease had relevant alterations of liver enzymes and coagulative profile, probably due to the innate immune response against the virus. Further studies are needed to better investigate the causes of liver injury in patients with COVID-19 disease and the effect of treatment for COVID-19 on the liver.

The novel coronavirus disease 2019 (COVID-19) is becoming a global pandemic with a total of approximately 420 000 confirmed cases, nearly 19 000 confirmed deaths and 197 countries, areas or territories involved.[1] These numbers were updated to March 26, 2020.[1] Worldwide, at present, chronic liver disease is a relevant disease burden.[2–4] In particular, the most common causes of liver diseases are chronic viral hepatitis, non-alcoholic fatty liver disease and alcohol.[2–4] Given this premises, investigating the influence of chronic liver disease on the novel COVID-19 is of scientific interest. In fact, accumulating evidence now suggests that patients with severe COVID-19 disease may have higher serum aminotransferase levels when compared to those with non-severe COVID-19 disease and, in some cases, even a significant liver injury.[3,4] However, to date, there is still scarce information regarding the global prevalence of chronic liver disease in individuals with COVID-19 disease. Therefore, aim of this study was to assess the overall prevalence of chronic liver disease among patients with COVID-19 disease by meta-analysing data in the observational studies available so far and to investigate the potential relationship between liver damage and COVID-19.

Studies were selected if they had information on history of chronic liver disease, as well as data on serum liver enzymes (ie aspartate aminotransferase [AST], alanine aminotransferase [ALT]). No restrictions in terms of race or ethnicity were adopted. Exclusion criteria were: (a) abstracts, reviews, editorials, guidelines; and (b) studies conducted in paediatric populations. Eleven observational studies[5–15] were identified by systematically searching PubMed (which has a specific section regarding COVID-19 disease) from January 1, 2020 to March 24, 2020 (date last searched) using the free text terms "COVID19" (OR "COVID-19" OR "Coronavirus disease 2019"). Non-English-language papers were excluded. Two investigators (AM and AD) independently examined the titles and abstracts and obtained full texts of potentially relevant papers. Working independently, we read the papers and determined whether they met inclusion criteria. Discrepancies were resolved by discussion with a third person (GB). For all eligible studies,[5–15] we extracted information regarding study size, source of data, population characteristics and outcome of interest. Data from eligible studies were extracted and meta-analysis was performed using random-effects modelling. All statistical tests were two sided and used a significance level of P < .05. We used STATA® 14.2 (Stata) for all statistical analyses. Specifically, metaprop command was used for random effect meta-analyses. We used the Score (Wilson) method to compute the confidence intervals.

Table 1 shows the main clinical characteristics of eligible studies. Figure S1 reports the flow diagram of the literature research and study selection. We included eleven observational studies[5–15] for a total of 2034 (predominantly Chinese) adult individuals (median age 49 years [IQR 45–54], 57.2% [n = 1165] men); 62 of whom had a prior history of chronic liver disease. Ten studies involved Chinese individuals,[5–10,12–15] whereas one enrolled US patients.[11] Only the study of Arentz et al involved patients over 65 years.[11] When reported, the main cause of chronic liver disease was attributed to the infection of hepatic B virus (HBV) or hepatic C virus (HCV). The prevalence of chronic liver disease among eligible case studies is plotted in Figure 1. As shown, the overall prevalence of chronic liver disease at baseline was relatively low, being of 3% (95% CI 2%-4%; I2 = 29.1%). Figure S2 shows the funnel plot of standard error by prevalence of chronic liver disease. The Egger's regression test (P = .015) showed statistically significant asymmetry for the funnel plot, thus suggesting a potential publication bias.[16] Subsequently, we used the non-parametric trim and fill analysis,[17] indicating that the impact of publication bias on random effect overall proportion was little (Table S1).

Figure 1.

Forest plot and meta-analysis of prevalences of chronic liver disease in adult patients with COVID-19

Among eligible studies, at baseline, mild liver test alterations were observed in most patients with COVID-19, even before the use of various medications (Table 1). For instance, regarding serum AST levels the minimum value was 26 IU/L and the maximum value was 46 IU/L. With regard to serum ALT levels the minimum value was 22 IU/L and the maximum value was 41 IU/L. When available, similar considerations can be made for total bilirubin levels and values of prothrombin time (Table 1). It is important to note that when liver function tests for patients with different durations of symptoms were assessed, later presentation was not consistently associated with important alterations of liver function tests.[3,4,14] However, when a specific comparison between patients with non-severe COVID-19 disease and those with severe COVID-19 disease was performed,[6,7,9,10,13–15] it was possible to note that patients with severe COVID-19 disease tended to have higher levels of liver enzymes, as well as a greater activation of coagulative and fibrinolytic pathways. For instance, in a recent study of nearly 1100 Chinese patients, Guan et al documented that elevated serum AST levels were observed in nearly 18% of patients with non-severe COVID-19 disease and in approximately 56% of patients with severe COVID-19 disease.[15] Moreover, in that study, elevated serum levels of ALT were also observed in nearly 20% of patients with non-severe COVID-19 disease and in approximately 28% of patients with severe COVID disease.[15] Similar findings were also observed in the study of Huang et al,[13] where the authors found that patients with severe COVID-19 disease had increased incidence of abnormal liver function.

These observations are in line with the notion that, along with other respiratory viruses that are able to determinate elevations of liver function biomarkers, this novel coronavirus may produce, in some cases, a relevant hepatic damage, probably through the immune interactions requiring the action of intrahepatic cytotoxic T cells and Kupffer cells.[2,4] Therefore, seeing the relatively low prevalence of chronic liver disease at baseline in patients with COVID-19, along with other authors,[2–4] we suppose that the liver damage observed in those with severe COVID-19 disease is essentially due to a dysregulated innate immune response against the virus rather than to the presence of specific and severe underlying liver diseases. Other coexisting causes for abnormal liver function in patients with severe COVID-19 disease may be the use of hepatotoxic drugs (eg acetaminophen) and the development of a systemic inflammatory response or multiple organ dysfunction.[2–4]

Our study has some important limitations that should be mentioned. First, the information regarding the prior history of chronic liver disease is not consistently reported in all observational studies on COVID-19 available so far, thereby limiting our ability to select a relevant number of studies. Second, information regarding the causes of chronic liver disease was not available for all eligible observational studies. Third, the individuals included in this study were mainly Chinese and were also relatively young. Therefore, the generalisation of our findings to other patient populations should be made with caution.

In conclusion, our study shows that at baseline the prevalence of chronic liver disease is relatively low in patients with COVID-19. However, individuals with severe forms of COVID-19 tend to develop important alterations of liver enzymes and to have changes of coagulative and fibrinolytic pathway profile, due to the innate immune response against the virus. Further studies are needed to better investigate the causes of liver injury in patients with COVID-19 and the effect of treatment for COVID-19 on the liver.

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....